About the company
When you are recruiting for a recognizable and respected brand...
... there are many people who want to build their professional careers with you.
That was (and still is) the case with Kinguin. Kinguin is a marketplace for digital products - it allows sellers from around the world to list offers for game keys, software licenses, gift cards, and other digital products.
There is no shortage of people eager to join this fantastic team.
Problem
Customer Care positions – which can attract up to several hundred applications within the first week after posting the job
To join the Customer Care team at Kinguin, you need to be available to work shifts 24/7, have a high level of English (nice to have: experience working in English) and experience in online support.
In some locations, there are so many applicants that the job ad has to be taken down just a few days after publication. The Kinguin team reviews them on an ongoing basis to ensure that candidates are processed efficiently and in accordance with the highest standards of candidate experience.
Goal
Efficient and accurate application preselection without compromising the candidate experience
Managers at Kinguin wanted to streamline processes, but did not want to implement solutions that would exclude applicants based on keywords. There are signs on the market of such solutions, which raise legitimate concerns among candidates.
The Kinguin team was open to supporting the pre-selection of candidates with the help of AI, but without giving up the human touch and reviewing the content of candidates' resumes.
TRAFFIT & its role
Candidate scoring – real support for pre-selection based on criteria specified by the recruiter
It turned out that this need is perfectly met by the AI-powered candidate scoring feature available in Traffit. The ATS suggests job match criteria based on the information provided about the position and the hiring manager's requirements - the person processing applicants can freely adjust them.
Each new application is evaluated (on a scale of 0-100%) according to these criteria, along with a justification for the evaluation and a transparent list of criteria that have and have not been met.
However, these concerns were quickly dispelled - the fact that the model is based on the context of the resume, rather than just keywords, changes everything!”
Along with the assessment itself, Kinguin's team also gained justification for each score and follow-up questions with a list of topics to be discussed during recruitment interviews. These are personalized and justified from the perspective of better understanding the candidate's suitability for the role.
The scoring mechanism is “aware” that the absence of certain information in a resume does not mean that given criteria has not been met.
Results
Candidates selected with AI with higher language test pass rate
The Kinguin team has confirmed that scoring candidates in Traffit made it much easier for them to select suitable profiles from hundreds of applications. An example of this is the recent recruitment for Customer Care department in the Philippines, during which the Kinguin team received several hundred applications. This was the perfect opportunity to showcase candidate scoring with the help of AI.
In this type of vacancy, selected candidates take an English language test. Typically, about 40-50% of selected applicants pass this test. Among those selected for the test with the help of scoring, the pass rate was as high as 80%
Their suitability for the position was later confirmed by subsequent stages of the recruitment process!
In this recruitment process, Kinguin's recruitment team was joined by a new member who deals with internal human resources on a daily basis.
Thanks to the support of AI scoring, her effectiveness in pre-selecting candidates to move on to the English language verification stage was just as good as that of experienced recruiters!
To pass the test, you must score at least 70% of the possible points.
The pass rate among candidates selected for the AI-assisted test was
higher than usual, at 80% compared to the standard 50%"
Conclusions
The value of the score generated by AI is directly proportional to how well the matching criteria have been selected
The model will work as well as we communicate our needs to it. Writing criteria is a skill that must be learned—sometimes through trial and errors. But thanks to the transparent justification of each score, the recruiter can immediately see that the criteria have been selected correctly (or need to be changed).
The story of Kinguin and recruitment with the support of candidate scoring is a great example of how AI technology can go hand in hand with recruiters, providing them with real support. On its own terms.